INDEX

Explanations

expressions of comparison and expectations about quality or performance

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

opper

-0.17

').'

-0.16

endid

-0.15

ÅĻÃŃd

-0.15

agar

-0.14

 DÃ¼n

-0.14

IRROR

-0.14

abis

-0.14

ulton

-0.14

Ð¾ÑģÑĤÐ°ÑĤ

-0.14

POSITIVE LOGITS

ãĤĤãģ£ãģ¨

0.21

 ideal

0.20

 could

0.17

 missing

0.17

 flash

0.17

0.16

 desired

0.16

 wanted

0.15

ideal

0.15

 wish

0.15

Activations Density 0.193%

expressions of comparison and expectations about quality or performance

No Comments

No Known Activations